UA120372C2 - Decoding method and decoder for dialog enhancement - Google Patents

Abstract

There is provided a method for enhancing dialog in a decoder of an audio system. The method comprises receiving a plurality of downmix signals being a downmix of a larger plurality of channels; receiving parameters for dialog enhancement being defined with respect to a subset of the plurality of channels that is downmixed into a subset of the plurality of downmix signals; upmixing the subset of downmix signals parametrically in order to reconstruct the subset of the plurality of channels with respect to which the parameters for dialog enhancement are defined; applying dialog enhancement to the subset of the plurality of channels with respect to which the parameters for dialog enhancement are defined using the parameters for dialog enhancement to provide at least one dialog enhanced signal; and subjecting the at least one dialog enhanced signal to mixing to provide dialog enhanced versions of the subset of downmix signals.

Description

The field of technology

The invention disclosed herein relates generally to audio coding.

In particular, it relates to methods and devices for performing dialogue amplification in channel-based audio systems.

Prerequisites of the invention

Dialogue amplification consists in performing dialogue amplification in relation to other audio content. It can be used, for example, to enable hearing-impaired people to follow dialogue in a film. For channel-based audio content, dialogue is typically present in multiple channels and mixed with other audio content. Therefore, strengthening the dialogue is a non-trivial task.

There are several ways to perform dialogue enhancement in the decoder. Some of these methods first decode the full channel content, i.e. the full channel configuration, and then use the resulting dialogue gain parameters to predict dialogue based on this full channel content. The predicted dialogue is then used to amplify the dialogue in the appropriate channels. However, such decoding methods rely on a decoder capable of decoding the full channel configuration.

However, decoders with low complexity, as a rule, are not designed to decode the full configuration of channels. Instead, a low-complexity decoder can decode and output a smaller number of channels that represent a down-mixed version of the full channel configuration. Accordingly, in a low-complexity decoder, the full channel configuration is not available. Since the dialogue enhancement parameters are defined with respect to the channels of the full channel configuration (or at least with respect to some of the channels of the full channel configuration), the known dialogue enhancement methods cannot be directly applied by a low-complexity decoder.

In particular, this is due to the fact that the channels to which the dialogue enhancement parameters are applicable can still be mixed with other channels.

Thus, there is room for improvements that allow a low-complexity decoder to be used to enhance dialogue without having to decode the full channel configuration.

Brief description of graphic materials

In the following, the exemplary embodiments will be described in more detail with reference to the accompanying graphic materials, in which: fig. 1 is a schematic illustration of a channel configuration 7.144 downmixed into a downmix configuration 5.1 according to a first downmix scheme; fig. 10 is a schematic illustration of channel configuration 7.144 downmixed into downmix configuration 5.1 according to the second downmix scheme; Fig. 2 is a schematic illustration of a prior art decoder for performing dialogue amplification on a fully decoded channel configuration; fig. 3 -- a schematic illustration of dialogue enhancement according to the first method; fig. 4 -- a schematic illustration of dialogue enhancement according to the second method; fig. 5-- a schematic illustration of the decoder according to the examples given for implementation; fig. 6-- a schematic illustration of the decoder according to the examples given for implementation; fig. 7 - a schematic illustration of the decoder according to the examples given for implementation; fig. 8 is a schematic illustration of an encoder corresponding to any of the decoders shown in FIG. 2, fig. 5, fig. 6 and fig. 7;

BO fig. 9 -- diagrams of methods for calculating operation BA of combined data processing, which consists of two sub-operations A and B, based on the parameters that control each of these sub-operations.

All figures are schematic and generally show only those elements that are necessary to illustrate the present invention, while other elements may be omitted or merely suggested.

Detailed description of the invention

In view of the above, it is an object of the present invention to provide a decoder and related methods that allow the use of dialogue enhancement without the need to decode a complete channel configuration.

I. Review

According to the first feature, the examples of implementation provide a way to enhance the dialogue in the decoder of the audio system. This method includes the following steps: receiving a number of down-mixing signals, which are the result of down-mixing a larger number of channels; receiving parameters to enhance dialogue, and these parameters are defined with respect to a subset of the channel array that includes channels that contain dialogue, and this subset of the channel array is downmixed into a subset of the downmix signal array; adoption of recovery parameters, which provide the possibility of parametric recovery of downmixed channels into a subset of downmixed signals; parametric up-mixing of a subset of a number of signals; applying dialog enhancement to a subset of the channel array for which the dialog enhancement parameters are defined, using the dialog enhancement parameters to provide at least one dialog enhanced signal; and downmixing the at least one dialogue-enhanced signal to provide dialogue-enhanced versions of a subset of the plurality of downmixed signals.

With such a scheme, the decoder does not have to restore the full channel configuration to perform dialogue amplification, which reduces complexity. Instead, the decoder restores those channels that are needed to apply dialogue enhancement. This includes, in particular, a subset of a number of channels, in relation to which the accepted parameters for enhancing the dialogue are defined.

After dialog enhancement is performed, that is, when at least one dialog-enhanced signal has been determined based on the dialog-enhanced parameters and the subset of the channel array for which those parameters are defined, the dialog-enhanced versions of the received downmix signals are determined by subjecting the dialog-enhanced signal(s) to mixing procedure. As a result, dialogue-enhanced versions of the downmix signals are obtained for subsequent reproduction by the audio system.

In the exemplary embodiments, the upmixing operation may be full (one that restores the entire set of encoded channels) or partial (one that restores a subset of the channels).

For the purposes of this document, a downmix signal refers to a signal that is a combination of one or more signals/channels.

For the purposes of this document, parametric upmixing refers to the recovery of one or more signals/channels from a downmixed signal using parametric techniques. It should be emphasized that the exemplary embodiments disclosed herein are not limited to channel-based content (in the sense of audio signals associated with fixed or predetermined directions, angles, and/or positions in space), but also extend to object-based content.

According to the options given for example, at the stage of parametric up-mixing of a subset of down-mixing signals, the decorrelated signals are not used for the purpose of restoring the subset of the channel series, for which the parameters for dialogue amplification are defined.

The advantage of this is that the computational complexity is reduced, while simultaneously increasing the quality of the resulting dialogue-enhanced versions of the downmix signals (ie, output quality). In more detail, the benefits obtained by using decorrelated signals in upmixing are reduced in the next downmix to which the signal is subjected to enhanced dialogue. Because of this, the use of decorrelated signals can be mostly avoided, thereby reducing the complexity of the calculations. In fact, using decorrelated signals in upmixing in conjunction with dialogue amplification could result in poorer quality as it could cause the decorrelator to reverberate on the amplified dialogue.

According to the exemplary embodiments, the mixing is performed according to the mixing parameters that describe the contribution of at least one dialogue-enhanced signal to the dialogue-enhanced version of a subset of a number of downmix signals.

Thus, there may be certain mixing parameters that describe how at least one up-mixed signal should be mixed in order to provide up-dialog versions of a subset of a number of downmixed signals. For example, the downmix parameters may be in the form of weights that describe how much at least one upmixed signal should be downmixed into each of the downmixed signals in a subset of the downmixed signals to produce the downmixed versions of the subset of downmixed signals. Such weights may, for example, take the form of representation parameters that are indicative of the spatial positions associated with at least one cross-talk signal relative to the spatial positions associated with a number of channels and thus with a corresponding subset of the downmix signals. According to other examples, the mixing parameters may specify whether or not at least one up-dialog signal should contribute to, eg, be part of, one particular up-dialog version of a subset of the downmix signals. For example, "1" may indicate that the dialogue-enhanced signal should be included when forming one particular dialogue-enhanced version of the downmix signals, and "0" may indicate that it should not be included.

In the step of downmixing at least one enhanced dialog signal to provide enhanced dialog versions of a subset of the downmix signals, the enhanced dialog signals may be mixed with other signals/channels.

According to the exemplary embodiments, at least one dialogue-enhanced signal is mixed with channels that are recovered in the up-mixing step, but which have not been dialogue-enhanced. In more detail, the step of parametrically up-mixing a subset of the down-mixing signals may include restoring at least one additional channel in addition to the number of channels for which dialog enhancement parameters are defined, wherein the mixing includes mixing the at least one additional channel together with the at least one enhanced dialog signal. For example, all downmixed channels in a subset of the downmix signal can be restored and mixed. In such embodiments, there is typically a direct correspondence between each enhanced dialogue signal and a specific channel.

According to other exemplary embodiments, at least one signal with enhanced dialogue is mixed with a subset of a number of downmixed signals. More

In more detail, the step of parametrically upmixing a subset of the downmix signals may include recovering only the subset of the channel array for which the dialog enhancement parameters are defined, and the step of applying dialog enhancement may include predicting and amplifying the dialog component from the subset of the channel array for which the parameters are defined for enhancing dialogue, using parameters for enhancing dialogue to provide at least one enhanced dialogue signal, and mixing may include mixing at least one enhanced dialogue signal with a subset of the downmix signals. Thus, such implementations serve to predict and amplify dialogue content and mix it into a subset of downmix signals.

In general, it is worth noting that a channel can contain dialog content mixed with non-dialog content. In addition, dialog content that corresponds to one dialog can be mixed into multiple channels. The prediction of a dialogue component from a subset of a number of channels, in relation to which the parameters for strengthening the dialogue are defined, is usually understood as the fact that the dialogue content is obtained, that is, extracted, from the channels and combined in order to restore the dialogue.

The quality of dialogue amplification can be further enhanced by accepting and using the audio signal that represents the dialogue. For example, an audio signal representing dialogue may be encoded at a low bit rate, resulting in audible artifacts when listening to it separately. However, when used together with parametric dialogue enhancement, i.e. when using dialogue enhancement parameters at the stage of applying dialogue enhancement to a subset of a number of channels for which dialogue enhancement parameters are defined, the resulting dialogue enhancement can be improved, for example, in terms of sound quality. In more detail, the method may further include receiving an audio signal representing the dialogue, wherein the step of applying the dialogue enhancement includes applying the dialogue enhancement to a subset of the number of channels for which the dialogue enhancement parameters are defined, with additional use of the audio signal representing the dialogue.

In some embodiments, the mixing parameters may already be available in the decoder, for example, they may be hard-coded. In particular, this may be the case 60 when at least one enhanced dialogue signal is always mixed in the same way,

for example, if it is always mixed with the same restored channels. In other embodiments, the method includes receiving mixing parameters for the step of mixing at least one signal with enhanced dialogue. For example, mixing parameters can form part of dialog gain parameters.

According to the exemplary embodiments, the method includes receiving mixing parameters that describe a down-mixing scheme that describes which down-mix signal each of a number of channels is mixed into. For example, if each signal with enhanced dialogue corresponds to a channel, which, in turn, is mixed with other restored channels, then the mixing is carried out in accordance with the down-mixing scheme so that each channel is mixed into the correct down-mixing signal.

The downmixing scheme can change over time, i.e. it can be dynamic, thus increasing the flexibility of the system.

The method may also include receiving data that identifies a subset of the channel array for which dialog enhancement parameters are defined. For example, data that identifies a subset of the channel array for which the dialog enhancement parameters are defined may be included in the dialog enhancement parameters. In this way, the decoder can be given a signal about which channels should be enhanced with dialogue. Alternatively, such information may be available in the decoder, for example, it may be hard-coded, meaning that the dialogue enhancement parameters are always defined for the same channels. In particular, the method may also include receiving information indicating which signals from the enhanced dialogue signals should be mixed. For example, the method of this embodiment may be implemented by a decoding system operating in a particular mode in which the enhanced dialogue signals are not mixed back into a completely identical set of downmix signals that was used to provide the enhanced dialogue signals. Thus, the mixing operation in practice may be limited to an incomplete sample (one or more signals) from a subset of the downmix signal range. Other dialogue-enhanced signals add to slightly different downmix signals, such as downmix signals that have undergone format conversion. As soon as the data identifying the subset of the channel array for which the parameters for dialog enhancement are defined and the downmixing scheme are known, the subset of the downmix signal array that is downmixed with the subset of the channel array for which the parameters for dialog enhancement are defined is known. In more detail, the data identifying the subset of the channel array for which the dialog enhancement parameters are defined can be used in conjunction with the downmixing scheme to find the subset of the downmix signal array that is downmixed to the subset of the channel array for which the dialog enhancement parameters are defined .

The steps of up-mixing a subset of down-mixing signals, applying dialogue enhancement, and mixing can be performed as matrix operations defined, respectively, by restoration parameters, dialogue enhancement parameters, and mixing parameters. The advantage of this is that the method can be efficiently implemented by performing matrix multiplication.

Moreover, the method may include combining, by matrix multiplication, the matrix operations that correspond to the steps of up-mixing a subset of the down-mixing signal series, applying dialog gain and mixing, into a single matrix operation before applying the down-mixing signal to a subset of the series. Thus, various matrix operations can be combined into a single matrix operation, thus further increasing the efficiency and reducing the computational complexity of the method.

The dialog gain parameters and/or recovery parameters may be frequency dependent, thus allowing the specified parameters to be different for different frequency bands.

In this way, dialogue gain and restoration can be optimized in different frequency bands, thereby increasing the quality of the output sound.

In more detail, parameters for dialogue enhancement may be defined with respect to a first set of frequency bands, and restoration parameters may be defined with respect to a second set of frequency bands, wherein the second set of frequency bands is different from the first set of frequency bands.

This may be preferable when reducing the bitrate for transmitting dialog enhancement parameters and restoration parameters in the bitstream, when, for example, the restoration process requires parameters with a higher frequency resolution than the dialog enhancement process, and/or when, for example, the enhancement process the dialog is performed on a lower bandwidth than the recovery process.

According to the exemplary embodiments, the (predominantly discrete) values of the parameters for enhancing the dialog can be taken multiple times and associated with the first set of time points in which the corresponding values are applicable exactly. In this specification, a statement that a value is applicable, or known, "exactly" at a specified point in time shall mean that the value has been received by the decoder, usually together with an explicit or implicit indication of the point in time at which it is applicable In contrast, a value that is interpolated or predicted for a particular point in time is in this sense not applicable "exactly" at that point in time, but represents an estimate on the part of the decoder. "Exactly" does not imply that the given value achieves an exact reproduction of the audio signal. A predetermined first interpolation scheme may be established between successive time instants in the set. An interpolation scheme that determines how to estimate the approximate value of a parameter at a certain point in time that is between two extreme points in time in the set at which the parameter values are known can be, for example, linear or piecewise-constant interpolation. If the time point of prediction is at a certain distance from one of the extreme points of time, the linear interpolation scheme is based on the assumption that the value of the parameter at the time of prediction depends linearly on the specified distance, while the piecewise-constant interpolation scheme ensures that the value of the parameter does not varies between each known and subsequent value. Other possible interpolation schemes may also take place, including, for example, schemes in which polynomials with degree greater than unity, splines, rational functions, Gaussian processes, trigonometric polynomials, wavelets or their combination. The set of time instants may not be explicitly transmitted or declared, but instead may be inferred from an interpolation scheme, such as the start point or end point of a linear interpolation interval, which may be implicitly bound to the frame boundaries of the audio processing algorithm. The recovery parameters can be obtained in a similar way: the (mostly discrete) values of the recovery parameters can be associated with a second set of time points, and a second interpolation scheme can be performed between successive time points.

The method may also include selecting a type of parameters, wherein the parameters of the type are either dialog enhancement parameters or restoration parameters, such that the set of time points associated with the selected type contains at least one prediction point that is a time not in the set associated with the unselected type. For example, if the set of points in time to which the recovery parameters are associated contains a particular point in time that is not present in the set of points in time to which the parameters to enhance dialog are associated, then that particular point in time will be the point in time of the prediction if the parameters of the selected type there are recovery parameters, and the parameters of the unselected type are parameters to strengthen the dialogue. Similarly, in another situation, the prediction instant can instead be found in the set of instants to which the dialogue enhancement options are associated, and then the selected and unselected types are swapped. Preferably, the selected parameter type is the type that has the highest density of time points with associated parameter values, in this use case this may reduce the total number of prediction operations required.

Values of parameters of an unselected type at the time of prediction can be predicted.

The prediction can be performed using a suitable prediction method, such as interpolation or extrapolation, given a predefined interpolation scheme for parameter types.

The method may include a calculation step based on at least one predicted value of the parameters of the unselected type and the accepted value of the parameters of the selected type of the combined processing operation, which is at least up-mixing of a subset of down-mixing signals followed by amplification of the dialogue at the moment of prediction. In addition to the recovery parameter values and dialogue enhancement parameters, the calculation may be based on other values, such as downmix parameter values, and the combined processing operation may also represent a step of mixing the dialogue-enhanced signal back into the downmix signal.

The method may include the step of calculating based on at least the (accepted or predicted) value of the parameters of the selected type and at least the (accepted or predicted) value of the parameters of the unselected type, and at least one of the specified values represents the accepted value of the combined processing operation at the adjacent time point in to the set associated with the selected or unselected type. The adjacent point in time can be either earlier or more distant than the point in time of prediction, and the requirement that this adjacent point in time be the nearest neighbor in terms of distance is not essential.

In this method, the steps of up-mixing a subset of the down-mix signals and applying the dialog gain can be performed between the time of prediction and the adjacent time using the interpolated value of the calculated combined processing operation. With the help of interpolation of the calculated combined processing operation, a reduction in computational complexity can be achieved. Because the two types of parameters are not interpolated separately, and because they do not form a product (i.e., a combined processing operation), at each interpolation point, less may be required to achieve an equally useful result in terms of perceived listening quality. the number of mathematical operations of addition and multiplication.

According to additional exemplary embodiments, a combined processing operation at a contiguous point in time can be calculated based on the accepted value of the parameters of the selected type and the predicted value of the parameters of the unselected type. The reverse situation is also possible, in which the combined processing operation at the adjacent moment of time can be calculated on the basis of the predicted value of the parameters of the selected type and the accepted value of the parameters of the unselected type. Situations in which parameter values of the same type are the accepted value at the time of prediction and the predicted value at an adjacent time point may arise if, for example, the time points in the set to which the parameters of the selected type are associated are strictly between points in time time in the set to which parameters of the unselected type are associated.

According to the exemplary embodiments, the combined processing operation at a contiguous point in time can be calculated based on the accepted value of the parameters of the selected type of parameters and the accepted value of the parameters of the unselected type of parameters. Such situations may arise, for example, when the exact values of parameters of both types are accepted for the frame boundaries, but also -- for the selected type -- for the instant of time inside the boundaries. Then the contiguous instant of time is the instant of time associated with the frame boundary, and the instant of prediction time is in the middle between the frame boundaries.

According to additional exemplary embodiments, the method may also

To include making a selection based on the first and second interpolation schemes of the combined interpolation scheme according to a predetermined selection rule, wherein the interpolation of the respective calculated combined processing operations corresponds to the combined interpolation scheme. The predetermined selection rule may be determined for the case in which the first and second interpolation schemes are the same, and may also be determined for the case in which the first and second interpolation schemes are different. For example, if the first interpolation scheme is linear (its, preferably, if there is a linear relationship between the parameters and the quantitative properties of the dialogue amplification operation), and the second interpolation scheme is piecewise-continuous, then the combined interpolation scheme can be chosen linear.

According to the options given for the example, the prediction of the value of the parameters of the unselected type at the time of prediction is performed according to the interpolation scheme for the parameters of the unselected type. This may include using the exact value of the parameter of the unselected type at a point in time in the set associated with the unselected type that is contiguous to the time of prediction.

According to the exemplary embodiments, the combined processing operation is calculated as a single matrix operation and then applied to a subset of a number of downmixed signals. Predominantly, the stages of upmixing and application of dialog enhancement are performed as matrix operations, which are defined by restoration parameters and dialog enhancement parameters. A linear interpolation scheme can be selected as the combined interpolation scheme, and the interpolated value of the corresponding calculated combined processing operations can be calculated by linear matrix interpolation. To reduce the computational complexity, the interpolation can be limited to such matrix elements that change between the moment of prediction and the adjacent moment of time.

According to the exemplary embodiments, the received downmix signals can be divided into time frames, and the method in the steady state of operation can include the step of accepting at least one value of the parameters of the corresponding types that is exactly applicable at some point in time in each time frame. For the purposes of this document, "fixed mode" refers to work that does not include the presence of a beginning and end part, such as a song, and work that does not include internal transitions that make frame subdivision necessary.

According to the second feature, a computer software product is provided, which contains a machine-readable medium with commands for performing the method according to the first feature.

The machine-readable medium may be a permanent machine-readable medium or device.

According to the third feature, a decoder is provided for enhancing dialogue in an audio system, which includes: a receiving component configured to receive: a series of down-mixing signals, which are the result of down-mixing a larger number of channels, parameters for enhancing dialogue, and these parameters are defined with respect to a subset of the series of channels, which includes channels that contain dialogue, and this subset of the number of channels is subjected to downmixing into a subset of the number of downmixing signals, and restoration parameters that provide the possibility of parametric restoration of the channels subjected to downmixing into a subset of the number of downmixing signals; an upmixing component, made with the possibility of parametric upmixing of a subset of a number of downmixing signals based on restoration parameters in order to restore a subset of a number of channels for which the parameters for dialogue amplification are defined; and a dialog enhancement component configured to apply dialog enhancement to a subset of a number of channels for which dialog enhancement parameters are defined, using dialog enhancement parameters to provide at least one dialog enhanced signal; and a mixing component configured to mix at least one dialogue-enhanced signal to provide dialogue-enhanced versions of a subset of a number of downmix signals.

In general, the second and third features may have the same characteristics and advantages as the first feature.

II. Examples of implementation options are provided

In fig. and fig. 16 schematically presents the configuration of channels 7.1-4 (corresponding to the configuration of speakers 7.14-44) with three front channels, H, C, K, two surround channels, I 5, K5, two rear channels, IV, KV, four height channels TE, TEK,

TV, TVE, and the IEE channel of low-frequency effects. In the process of encoding the configuration of the 7.144 channels, as a rule, they are downmixed, that is, they are combined into a smaller number of signals, which are called downmix signals. In the downmix process, channels can be combined in different ways to create different downmix configurations. In fig. and the first configuration 100a of downmixing 5.1 with downmixing signals Y, c, r, I5, r5, Me is presented. The circles on the figure show which channels are downmixed into which downmix signals. In fig. 165 presents the second configuration 1006 of downmixing 5.1 with downmixing signals I, c, r, !I, her, Me. The second configuration is 100 rubles. downmixing 5.1 differs from the first configuration 100a of downmixing 5.1 in that the channels are combined in a different way. For example, in the first down-mixing configuration 100a, the HER and TEG channels are down-mixed into a signal | down-mixing, while in the second configuration 100U6 down-mixing down-mixing into the signal | channels I, 1/5, IV are subjected to down-mixing. The down-mix configuration in this document is sometimes referred to as a down-mix scheme, which describes which channels are down-mixed into which down-mix signals. The down-mixing configuration, or down-mixing scheme, can be dynamic in that it can vary between time frames of the audio coding system. For example, in some time frames the first down-mixing scheme 100a may be used, while in other time frames the second down-mixing scheme 1006 may be used. In the case of a dynamic change of the down-mixing scheme, the encoder can send data to the decoder indicating which down-mixing scheme was used when encoding the channels.

In fig. 2 illustrates a prior art dialogue enhancement decoder 200. This decoder contains three main components: a receiving component 202, an upmixing component 204, or recovery, and a dialogue enhancement (OE) component 206.

Decoder 200 is of the type that receives a series of downmix signals 212 , reconstructs the full channel configuration 218 based on the received downmix signals 212 , performs dialogue amplification on the full channel configuration 218 , or at least a subset thereof, and outputs the full channel configuration 220 with enhanced dialogue

In more detail, the receiving component 202 is configured to receive a stream of data (sometimes referred to as a bit stream) from the encoder 210 . The data stream 210 may contain data of various types, and the receiving component 202 may decode the received data stream 210 into data of various types. In this case, the data stream contains a series of downmix signals 212, restoration parameters 214, and dialog enhancement parameters 216.

The upmix component 204 then restores the full channel configuration based on a number of downmix signals 212 and recovery parameters 214 . In other words, the upmix component 204 restores all channels 218 that have been downmixed into downmix signals 212 . For example, the parameter-based upmixing component 204 of recovery 214 may parametrically recover the full channel configuration.

In the presented example, the down-mixing signals 212 correspond to the down-mixing signals of one of the down-mixing configurations 5.1 shown in FIG. Yes and 1p, and channels 218 correspond to the channels of channel configuration 7.14, which is presented in fig. Yes and 15. However, it will be understood that the principles of decoder 200 are applicable to other channel/downmix configurations.

The reconstructed channels 218 , or at least a subset of the reconstructed channels 218 , are then subjected to dialog enhancement using the dialog enhancement component 206 . For example, the dialog enhancement component 206 may perform a matrix operation on the reconstructed channels 218 or at least on a subset of the reconstructed channels 218 to produce dialogue enhanced channels. Such a matrix operation, as a rule, is defined by the parameters 216 of dialogue enhancement.

For example, the dialogue enhancement component 206 can subject C channels to dialogue amplification,

I, K in order to create channels Spr, I oye, Koe with enhanced dialogue, while the other channels are simply skipped, as shown in fig. 2 dotted lines. In such a situation, the dialogue amplification parameters are defined only in relation to channels C, I, EK, i.e. in relation to a subset of the number of channels 218.

For example, the dialogue enhancement parameters 216 can define a 3x3 matrix that can be applied to channels C, I, and B. AND

Ave tlia izz LbPza V

Alternatively, channels not involved in dialogue gain may be omitted using a dialogue gain matrix with "1" in the corresponding diagonal positions and "0" in all other elements of the corresponding rows and columns.

Svv pi 27" 0 ot. 000080 o.i

Tov toi yo 0 ota 0 000000 i,

TE, p 011 0000080 o ORI TC

Kk pz yza bot 00000 o0v6o Kk

TiEV on 00 o0 10000000 EC 51 o p 00 01000000 5 t vi. o 0 00001000 o0 ov ate 0 p EV. and about vaa1o00o0o05 yin

KU Ii p o o o 0 pop 1o0009 KU

TE at 00 00000010 OR TEBK

KE 0 0 EV. 0 o vyaooo1o KV

EE pap pao0opopaoi1 KBE

The dialogue enhancement component 206 may perform dialogue enhancement according to various methods. The first method, which in this document is called channel-independent parametric amplification, is presented in fig. 3. Dialogue amplification is performed with respect to at least a subset of restored channels 218, as a rule, channels that contain dialogue, here - channels I, E, C. Parameters 216 for dialogue amplification include a set of parameters for each of the channels to be amplified. In the presented example, the sets of parameters are represented by the parameters рі, р», рз, which refer, respectively, to channels Г, К, С. In principle, the parameters transmitted in this method represent the relative contribution of the dialog to the mixing energy for the frequency-time mosaic in channels In addition, the amplification factor d takes part in the process of dialogue amplification. The amplification factor can be expressed as: t d -10:0-1 where b is the amplification factor when amplifying the dialogue, expressed in dB. The gain factor for dialogue amplification may, for example, be input by the user, and therefore is typically not included in the data stream 210 shown in FIG. 2.

In the method of channel-independent parametric amplification, the dialog amplification component 206 multiplies each channel by its corresponding parameter p; and on the ud gain factor, and then adds the result to the channel, creating 220 channels with enhanced dialogue, here -- І ов,

Howl, Owl. Using matrix representation, this can be written as follows:

Хе - У ваша) milestone where А is a matrix that contains channels 2181, КЕ, С), ХЕ as terms. the matrix, which contains as lines 220 channels with enhanced dialogue, is a vector-line with elements corresponding to . with gave dialogue amplification parameters Ri, M, 3 for each channel, and aivdiri-. diagonal matrix, which contains elements ? on the diagonal.

The second method of dialogue enhancement, which in this document is called multi-channel dialogue prediction, is presented in fig. 4. In this method, the dialogue enhancement component 206 combines multiple channels 218 into a linear combination to perform prediction of the dialogue signal 419. In addition to coherently adding the presence of dialogue across multiple channels, this approach can benefit from subtracting background noise in a channel containing dialogue using another channel without dialogue. For this purpose, the dialog gain parameters 216 contain for each channel 218 a parameter that determines the coefficient of the corresponding channel when creating a linear combination. In the presented example, the dialogue amplification parameters 216 contain the parameters рі, рг2, рз, which relate, respectively, to channels І, К, С. As a rule, to generate prediction parameters on the encoder side, optimization algorithms of the minimum root mean square error (MM5E) are used.

The dialogue enhancement component 206 may then amplify, ie increase, the predicted dialogue signal 419 by using a gain factor d and add the dialogue enhanced signal to channels 218 to create dialogue enhanced channels 220. To add a signal with enhanced dialogue to the correct channels in the correct spatial position (otherwise it will not enhance the dialogue with the expected gain), panning between the three channels is transmitted using representation coefficients, here - n, r, Iz. Provided that the coefficients of the representation are such that they conserve energy, that is, the third coefficient of the representation can be determined from the first two coefficients so that:

ІІ and . and B 11-76 - KI, with che -

Using a matrix representation of the dialogue gain performed by the dialogue gain component 206 in the multi-channel dialogue prediction method can be written as follows:

Ha NURIEH or 1 caste ri potri rz

Xe Vtr 1vd'str r tr "X tp t pot ve 1 vdstatri where I is the matrix of the identical transformation, X-- the matrix that contains as terms the channels 218 (I, A, C), Xe-- the matrix that contains as rows, 220 channels with enhanced dialogue, -- a row vector with elements that correspond to the parameters of Рi, B, 3 dialogue amplification for each channel, 77 -- a column vector that contains as elements the coefficients 73, 75, 75 of the representation , and 4 is the amplification factor, and 5 p is 1022-11.

According to the third method, which is referred to in this document as a signal-parametric hybrid, the dialogue enhancement component 206 may combine any of the first and second methods with the transmission of an additional audio signal (wave signal) that represents the dialogue. The latter, as a rule, is encoded with a low bit rate, which leads to the appearance of clearly audible artifacts when listening to it separately. Depending on the properties of the signals of the channels 218 and dialogue, and on the bit rate of data transmission intended for coding the wave signal of the dialogue, the coder also determines the mixing parameter, Us, which indicates how the gain contributions should be divided between the parametric contribution (from the first or second method) and an additional sound signal that represents the dialogue.

In combination with the second method of strengthening the dialogue in the third method, it can be written as follows:

Ha -N'shsach YAN dos RIH or 1 dastz tri Yagotib oh 82 ti Vr Vz1sti u

Hvt | Yaz' iz ' ri 1 star» I sTz' vz shchi t y y

Zast YazstasT In 1 dzstz va di stz ya de yg-- is an additional sound signal that represents a dialogue, and a and dia - as (1026 - 1), az - («y (10 - 1).

For combination with channel-independent gain (first method) sound signal

I her g. i i: t

Yasya, which represents the dialogue, is accepted for each channel 218. In the hbed recording, the amplification of the dialogue can be written as follows:

He - di Vo in yiad(ri: dz) Kh.

In fig. 5 shows a decoder 500 according to the exemplary embodiments. Decoder 500 is of the type that decodes a series of down-mixed signals, which are the result of down-mixing multiple channels, for subsequent playback. In other words, the decoder 500 is different from the decoder shown in FIG. 2, by the fact that it is not made with the possibility of restoring the full configuration of the channels.

The decoder 500 includes a receiving component 502 and a dialogue amplification unit 503, which includes a component 504 upmixing, a dialogue amplification component 506 and a mixing component 508.

As explained with reference to fig. 2, the reception component 502 receives the data stream 510 and decodes it into its components, in this case - into a series of signals 512 of the downlink

From mixing, which is the result of down-mixing of a larger number of channels (cf. Fig. 1 and 15), and parameters for strengthening 516 dialog. In some cases, the data stream 510 also contains data that reflects the mixing parameters 522 . For example, mixing parameters can form part of the dialog enhancement parameters. In other cases, the mixing parameters 522 are already available in the decoder 500, for example, they may be hard-programmed in the decoder 500. In other cases, the mixing parameters 522 are available for multiple sets of mixing parameters, and the data in the data stream 510 provides an indication of which set from these several sets of mixing parameters are used.

These dialog enhancement parameters 516 are typically defined with respect to a subset of a number of channels. Data that identifies a subset of the channel array for which dialog enhancement parameters are defined may be included in the received data stream 510 , for example, as part of the dialog enhancement parameters 516 . Alternatively, a subset of the channel array for which the dialog enhancement parameters are defined may be hard-coded into the decoder 500. For example, with reference to FIG. and, the parameters for dialogue amplification 516 can be determined with respect to channels Y, TEI, subjected to down-mixing into the down-mixing signal, channel C, which is contained in the down-mixing signal c, and channels K, TEK, subjected to down-mixing into the down-mixing signal r. C for the purpose of illustration, let's assume that dialogue is present only in channels I, C and K. It is worth noting that the parameters for strengthening 516 dialogue can be defined in relation to channels that contain dialogue, such as channels G, C, EK, but can also be defined in relation to channels , which do not contain dialogue, such as, in this example, TE, TEK channels. Thus, background noise in a channel that contains dialogue can, for example, be subtracted by using another channel without dialogue.

A subset of channels, in relation to which the parameters for enhancing the dialogue 516 are defined, are subjected to down-mixing into a subset 512a of a number of down-mixing signals 512. In the presented example, the subset 512a of the downmixing signals contains the downmixing signals c, I and d. This subset of the downmixing signals 512a is fed to the dialog amplification unit 503. A suitable subset 512a of the downmix signals can be found, for example, based on knowledge of the subset of the channel array for which the dialogue enhancement parameters are defined and the downmix scheme.

The upmixing component 514 uses parametric techniques known in the art to restore the downmixed channels to a subset of downmix signals 512a. Recovery is based on recovery options 514. In particular, the upmixing component 504 restores a subset of the number of channels for which the parameters for enhancing the dialogue 516 are defined. In some embodiments, the upmixing component 504 restores only a subset of the channel array for which the dialogue enhancement 516 parameters are defined. These exemplary embodiments will be described with reference to FIG. 7. In other embodiments, the upmixing component 504 restores at least one more channel in addition to the subset of the number of channels for which the dialog enhancement 516 parameters are defined. These exemplary embodiments will be described with reference to FIG. b.

Recovery parameters can be not only time-varying, but also frequency-dependent. For example, recovery parameters can take different values for different frequency bands. This, as a rule, improves the quality of restored channels.

As is known in the art, parametric upmixing can typically involve forming decorrelated signals from upmixed input signals, and parametrically reconstructs the signals based on the input signals and the decorrelated signals. (See, for example, the book "Zrayia! Atsaio Rgosezzipd: MRES Zyshtotspa apa

OiShieg Arriisaiope" by authors DUegoip Vgeeraai and SnNtgizui Raiiieg, ИЗВМ:978-9-470-03350-0). However, the upmixing component 504 mainly performs parametric upmixing

From mixing without using any such decorrelated signals. The advantages obtained when using decorrelated signals, in this case, are reduced by the subsequent downmixing performed by the mixing component 508. Therefore, the use of decorrelated signals can preferably be omitted by the upmixing component 504, thereby reducing the computational complexity. In fact, using decorrelated signals in upmixing with dialogue amplification would result in poorer quality, as it could cause the decorrelator to reverberate on the dialogue.

The dialog enhancement component 506 then applies the dialog enhancement to a subset of the channel array for which the dialog enhancement parameters 516 are defined, in order to obtain at least one dialog-enhanced signal. In some embodiments, the enhanced dialog signal corresponds to the enhanced dialog versions of a subset of the channel array for which dialog enhancement parameters 516 are defined. This will be explained in more detail below with reference to FIG. 6. In other embodiments, the enhanced dialog signal corresponds to a predicted and enhanced dialog component from a subset of the channel array for which dialog enhancement parameters 516 are defined. This will be explained in more detail below with reference to FIG. 7.

Similar to recovery parameters, dialogue enhancement parameters can change over time as well as with frequency. In more detail, parameters for dialogue enhancement can take different values for different frequency bands. The set of frequency bands for which the restoration parameters are defined can be different from the set of frequency bands for which the dialogue enhancement parameters are defined.

The mixing component 508 then mixes the at least one dialogue-enhanced signal to provide dialogue-enhanced versions 520 of the subset 512a of the downmixed signals. In the presented example of the version 520 with enhanced dialogue, the subsets 512a of the downmix signals are of the form owls, Ire, Goye, which correspond to the signals c, I, g of the downmix, respectively.

Mixing may be performed according to the mixing parameters 522 that describe the contribution of at least one dialogue-enhanced signal to the dialogue-enhanced version 520 of the downmix signal subset 512a. In some embodiments, see fig. 60b, the specified at least one signal with enhanced dialogue is mixed together with the channels that have been restored by the upmixing component 504. In these cases, the 522 mixing parameters may correspond to a down-mixing scheme, see fig. And 1r, which describes which of the 520 downmix signals with enhanced dialogue should be mixed into which each channel should be mixed.

In other embodiments, see fig. 7, the specified at least one enhanced dialogue signal is mixed together with a subset 512a of the downmix signals. In this case, the downmix parameters 522 may correspond to weights that describe how at least one dialogue-enhanced signal should be weighted in the subset 512a of the downmix signals.

The upmixing operation performed by the upmixing component 504 , the dialog boost operation performed by the dialog enhancement component 506 , and the mixing operation performed by the mixing component 508 are generally linear operations, each of which can be defined by a matrix operation, that is, by product of a matrix and a vector. This is true, at least, if the upmixing operation does not use decorrelator signals. In particular, the matrix associated with the upmixing operation is/can be obtained from the recovery parameters 514. In this regard, it should be noted that the use of decorrelator signals in the upmixing operation is ultimately possible, but the generation of the decorrelated signals is then not part of the matrix operations for upmixing. The upmixing operation with decorrelators can be thought of as a two-stage approach. In the first stage, the downmixing inputs are applied to the predecorrelator matrix, and each of the outputs after applying the predecorrelator matrix is applied to the decorrelator. In the second stage, the downmixing inputs and the output signals from the decorrelators are fed to an upmixing matrix, where the upmixing matrix coefficients corresponding to the downmixing inputs form what is called a "dry upmixing matrix", and coefficients corresponding to the outputs from the decorrelators form what is called a "wet upmix matrix". Each submatrix is mapped to a configuration of upmix channels. When the decorrelator signals are not used, the matrix associated with the upmixing operation is made with the ability to act only on the input signals 512a, and the columns that relate to the decorrelated signals (wet upmixing matrix) are not included in the matrix. In other words, the upmixing matrix in this case corresponds to the dry upmixing matrix. However, as noted above, the use of decorrelator signals in this case will, as a rule, lead to worse quality.

The matrix associated with the dialog gain operation (MU) is determined/can be obtained from the dialog gain parameters 516, and the matrix associated with the mixing operation is determined/can be obtained from the mixing parameters 522.

Since the upmixing operation, the dialog boosting operation, and the mixing operation are all linear operations, the corresponding matrices can be combined, by means of matrix multiplication, into a single matrix E (then ХрЕ ЕС Х and Э is АА, Here Х is a column vector downmix signals 512a, and Hvk is a column vector of dialogue-enhanced downmix signals 520. Thus, the entire dialogue-enhancement block 503 may correspond to a single matrix operation applied to a subset 512a of the down-mix signals to produce dialogue-enhanced versions 520 of said downmix signal subsets 512a Accordingly, the methods described herein can be implemented in an extremely efficient manner.

In fig. 6 shows a decoder 600 that corresponds to one of the exemplary embodiments of the decoder 500 shown in FIG. 5. Decoder 600 includes a receiving component 602 , an upmixing component 604 , a dialogue enhancement component 606 , and a mixing component 608 .

Similarly to the decoder 500 shown in fig. 5, the receiving component 602 receives the data stream 610 and decodes it into a series of downmix signals 612 , recovery parameters 614 , and dialog enhancement parameters 616 .

The upmix component 604 receives a subset 612a (corresponding to the subset 512a) of the downmix signal series 612. For each of the downmix signals in the subset 612a, the upmix component 604 restores all the channels that were downmixed in that downmix signal (Xe - CX). This includes channels 618a for which parameters are defined for dialogue enhancement, and channels 618B, which do not need to be included to enhance the dialogue. With reference to fig. 165, channels 6b18a, in relation to which the parameters for strengthening the dialogue are defined, can, for example, correspond to channels Г, И 5, С, К, К5, and channels 6186, which do not need to be included in strengthening the dialogue, can correspond to channels И В, КВ.

Channels 618a, in respect of which the parameters for dialogue enhancement C are defined, are then subject to dialogue enhancement using the dialogue enhancement component 606 (Y. - M. Hi, while channels 6186, which do not need to be included in the dialogue enhancement CX, bypass the dialogue enhancement component 606.

The dialog enhancement component 606 may use any of the first, second, and third dialog enhancement methods described above. In the case of using the third method, the data stream 610 may, as explained above, contain an audio signal that represents dialogue (ie, a coded wave signal that represents dialogue) to be used in amplifying the dialogue together with a subset 618a of the channel array for which parameters are defined for strengthening the dialogue

Still r.a y - j. (kh.-m TO

As a result, the dialogue enhancement component 606 outputs dialogue-enhanced signals 619, which in this case correspond to the dialogue-enhanced versions of the subset 618a of the channels for which the dialogue enhancement parameters are defined. For example, the enhanced dialog signals 619 may correspond to the enhanced dialog versions of channels I, I5, C, K, K5 shown in FIG. 16.

The mixing component 608 then mixes the enhanced dialogue signals 619 together with the Yew Heh ! | | (Xr C xg) channels 618r, which were not included in the dialogue enhancement "» with" in order to obtain versions 620 with enhanced dialogue of the subset 612a of downmixing signals. The mixing component 608 performs mixing according to the current down-mixing scheme, such as the down-mixing scheme shown in FIG. 15. In this case, the downmixing parameters 622 thus correspond to the downmixing scheme that describes which downmixing signal 620 each channel 619, 6186 should be mixed into. The downmixing scheme can be static and

This is known to the decoder 600, which means that the same down-mixing scheme is always used, or the down-mixing scheme can be dynamic, meaning that it can change from frame to frame, or it can be one of several schemes known in the decoder. In the latter case, the data stream 610 includes an indication of the down-mixing scheme.

In fig. 6, the decoder is equipped with an optional switching component 630. Switching component 630 can be used to transition between different downmixing schemes, for example, to transition from scheme 1006 to scheme 100ba. It is worth noting that the switching component 630, as a rule, leaves the signals c and Me unchanged, that is, with respect to these signals, it acts as a transit component. The switching component 630 may receive and act (not shown) on the basis of various parameters, such as, for example, recovery parameters 614 and parameters for enhancing dialogue 616 .

In fig. 7 shows a decoder 700 that corresponds to one of the exemplary embodiments of the decoder 500 shown in FIG. 5. Decoder 700 includes a receiving component 702 , an upmixing component 704 , a dialogue enhancement component 706 , and a mixing component 708 .

Similarly to the decoder 500 shown in fig. 5, the receiving component 702 receives the data stream 710 and decodes it into a series of down-mix signals 712 , recovery parameters 714 , and dialog amplification parameters 716 .

The upmix component 704 receives a subset 712a (corresponding to the subset 512a) of the downmix signal series 712. Unlike the embodiment described in relation to fig. b, the upmixing component 704 restores only the subset 718a of the range of channels for which the parameters for dialogue enhancement 716 are defined (in o. With reference to Fig. 160, the channels 718a for which the parameters for dialogue enhancement are defined may, for example, correspond to channels C, I, I 5, K, K5.

The dialogue enhancement component 706 then performs dialogue enhancement on the channels 718a for which the dialogue enhancement parameters are defined (Xa - Me: Xi In this case, the dialogue enhancement component 706 proceeds to predict the dialogue component based on the channels 7184 by forming a linear combination of the channels 718a according to the second by the dialog enhancement method. The coefficients used in forming this linear combination, denoted in Fig. 7 as ri---ro5, are contained in the dialog enhancement parameters 716. The predicted dialog component is then amplified by multiplying by a gain factor d to obtain a signal 719 with the amplification factor d can be expressed as:

Ge d - 1028-11 where sb is the amplification factor when amplifying the dialogue, expressed in dB. The gain factor for dialogue amplification may, for example, be entered by the user, and therefore it is typically not included in the data stream 710. It is worth noting that in the case where there are several dialog components, the above-described prediction and amplification procedure can be applied once for each dialog component.

The predicted dialogue-enhanced signal 719 (that is, the predicted and amplified dialogue components) is then mixed into a subset 712a of the downmix signals in order to obtain the dialog-enhanced versions 720 of the subset 712a of the downbox signals) mixing CE X / Mixing is performed with the parameters 722 of mixing, which according to describe the contribution of the dialog-enhanced signal 719 to the dialog-enhanced version 720 of the downmix signal subset. Mixing parameters are typically contained in the data stream 710. In this case, the mixing parameters 722 correspond to the weighting factors г", г2, из, which describe how at least one signal 719 with enhanced dialogue should be weighted in the subset 712a of the downmixing signals: т т- 100 ио

Hrv - X E "Xa - E 01 | m

ТZ Кк ) ше

Ka Ka 0 and 1 Hi

In more detail, the weights may correspond to representation factors that describe the panning of the at least one dialogue-enhanced signal 719 with respect to a subset 712a of the downmix signals, so that the dialogue-enhanced signal 719 is added to the downmix signals 712a in the correct spatial positions.

The representation coefficients (mixing parameters 722) in the data stream 710 may correspond to the upmixed channels 718a. In the presented example, there are five channels 718a subjected to upmixing, and thus there may be five corresponding representation coefficients gs1, gs2,..., po5. The values of I1, g2, gZ (corresponding to the downmix signals 712a) can then be calculated from gs1, gs2, . . . , go5 in conjunction with the downmix circuit. If multiple channels 718a correspond to the same downmix signal 712a, then the dialogue representation coefficients can be summed. For example, in the presented example, it turns out that г1:2гсС1, г2го2ягоЗ3, and г3-гс4-го5. In case the channels were downmixed using downmixing coefficients, this can also be a weighted summation.

It is worth noting that in this case, the dialog enhancement component 706 may also use an additional received audio signal that represents the dialog. In such a case, the predicted dialogue-enhanced signal 719 may be weighted together with the audio signal representing the dialogue before being fed to the mixing component 708

Sa - (1 - ai Ma Ha bo Appropriate weighting is provided by the Us Mixing parameter included in the parameters for dialogue amplification 716. The Us Mixing parameter indicates how the gain contributions should be divided between the intended dialogue component 719 (described above) and the additional audio signal that represents dialogue Os.

This is similar to what was described about the third method of dialogue amplification when it is combined with the second method of dialogue amplification.

In Fig. 7, the decoder is equipped with an optional switching component 730.

Switching component 730 may be used to transition between different downmixing schemes, for example, to transition from scheme 1006 to scheme 100a. It is worth noting that the switching component 730, as a rule, leaves the c and Me signals unchanged, that is, with respect to these signals, it acts as a transit component. The switching component 730 may receive and act (not shown) on the basis of various parameters, such as, for example, recovery parameters 714 and parameters for enhancing dialogue 716 .

The above was mainly explained in relation to channel configuration 7.14 and downmix configuration 5.1. However, it should be understood that the principles of decoders and decoding methods described in this document are equally applicable to other channel configurations and downmixing.

Fig. 8 is an illustration of an encoder 800 that can be used to encode a number of channels 818, some of which contain dialogue, to obtain a data stream 810 for transmission to the decoder. Encoder 800 may be used with any of decoders 200, 500, 600, 700. Encoder 800 includes a down-mixing component 805, a dialogue enhancement coding component 806, a parametric coding component 804, and a transmission component 802.

Encoder 800 receives a number of channels 818, for example, channels from configurations 100a, 1006 channels shown in FIG. But she is 165.

Downmixing component 805 performs downmixing of a number of channels 818 into a number of downmixing signals 812, which are then fed to transmission component 802 for inclusion in data stream 810. A number of channels 818 may, for example, be down-mixed according to a down-mixing scheme such as that shown in FIG. 1a or fig. 1 year

A number of channels 818 and downmix signals 812 are input to the parametric coding component 804. Based on its input signals, the parametric coding component 804 calculates recovery parameters 814 that enable recovery of channels 818 from downmix signals 812 . Recovery parameters 814 can be calculated, for example, using minimum mean square error (MMS) optimization algorithms as known in the art. The recovery parameters 814 are then fed to the transmission component 802 for inclusion in the data stream 810 .

The dialogue enhancement coding component 806 calculates parameters for dialogue enhancement 816 based on one or more of a number of channels 818 and one or more dialogue signals 813 . Dialog signals 813 represent dialog in its purest form. In particular, the dialog is already mixed into one or more channels 818. In the channels 818, thus, there may be one or more dialog components that correspond to the signals 813 of the dialog. Typically, the dialog enhancement coding component 806 calculates parameters for dialog enhancement 816 using least mean square error (MSE) optimization algorithms. These algorithms may provide parameters that enable the prediction of dialog signals 813 given some of the channel array 818. These dialog enhancement parameters 816 may thus be determined with respect to a subset of the channel array 818, namely those from which the signals 813 can be predicted. dialogue The parameters for prediction 816 of the dialog are submitted to the transmission component 802 for inclusion in the data stream 810.

Finally, the data stream 810 thus at least includes a number of down-mix signals 812 , recovery parameters 814 , and dialogue enhancement parameters 816 .

During normal decoder operation, parameter values of various types (such as dialog enhancement parameters or recovery parameters) are repeatedly received by the decoder at specific frequencies. If the frequencies at which the values of the various parameters are received are lower than the frequency at which the output from the decoder is to be calculated, then the values of the parameters may need to be interpolated. If it is known that the value of the generalized parameter P at points pi

Y2 in time is equal, respectively, to КУ213 and Ри: then the value ра) of this parameter at an intermediate moment of time of its BV Z 2 can be calculated using various interpolation schemes. One example of such a scheme, referred to herein as a linear interpolation scheme, can calculate an intermediate value using linear interpolation, such as piecewise-constant interpolation scheme, may instead involve keeping the parameter value tied to one of the known values throughout the time interval, e.g., РО) - c) or РК) - (22) or to a combination of known values, such as e.g. , the average value of v) - Gr) times LI, Information about which of the interpolation schemes should be used for a certain type of parameters during a certain period of time can be included in the decoder or provided to the decoder in various ways, as,

for example, together with the parameters themselves or as additional information contained in the received signal.

In one illustrative example, the decoder accepts parameter values for parameters of the first and second type. The accepted parameter values of each type are exactly applicable, respectively, in the first (T1-4И11, 72, M3, ...)3) and second (T2-421, 122, 123, ...)) sets of time moments, and the decoder also has access to information about how parameter values of each type should be interpolated in the event that the value needs to be evaluated at a point in time not present in the corresponding set. Parameter values control the quantitative properties of mathematical operations on signals, and these operations can, for example, be represented in the form of matrices. In the following example, it is assumed that the operation controlled by the parameters of the first type is represented by the first matrix A, the operation controlled by the parameters of the second type is represented by the second matrix B, and in this example the terms "operation" and "matrix" can be used interchangeably. At the time when it is necessary to calculate the output value from the decoder, it is necessary to calculate the combined operation of data processing, which corresponds to the composition of both operations. It is also assumed that matrix A represents an upmixing operation (controlled by restoration parameters) and that matrix B represents an application of dialog enhancement (controlled by parameters for dialog enhancement), and then, accordingly, the operation of upmixing processing followed by amplification is combined dialogue is represented by the matrix product BA.

Methods of calculating the combined processing operations are presented in fig. For--9e, where time passes along the horizontal axis, and the divisions of the axis indicate the moments of time at which it is necessary to calculate the combined processing operation (moments of output time). In the figures, triangles correspond to matrix A (representing the upmixing operation), circles to matrix B (representing the use of dialogue enhancement), and squares to matrices

BA of the combined operation (which represents the combined operation of upmixing followed by dialogue amplification). Colored triangles and circles indicate that the corresponding matrix is precisely known (that is, that the parameters governing the operation represented by the matrix are precisely known) at the corresponding instant in time, while uncolored triangles and circles indicate that the value of the corresponding matrix is predicted , or interpolated

Zo (for example, using any of the interpolation schemes described above). A colored square indicates that the BA matrix of the combined operation was calculated at the appropriate time, for example, by the matrix product of matrices A and B, and an uncolored square indicates that the BA value was interpolated from an earlier time point. In addition, the dashed arrows indicate between which time points the interpolation is performed. Finally, the solid horizontal line that connects the moments of time indicates that the values of the matrix in this interval are assumed to be piecewise constant.

In fig. 9a presents a method of calculating the combined operation of BA processing, which does not involve the use of this invention. The adopted values for operations A and B are exactly applicable at time points 111, 121 and 42, 122, respectively, and to calculate the matrix of the combined processing operation at each output time point, this method interpolates each of the matrices separately.

For each step forward in time, the matrix representing the combined processing operation is calculated as the product of the predicted values of A and B. It is assumed here that each matrix is to be interpolated using a linear interpolation scheme. If the matrix A contains M" rows and NO columns, and the matrix B contains M rows and A" columns, then each step forward in time would require OM) multiplication operations on each set of parameters (to perform the matrix multiplication necessary to calculate the matrix BA by combined processing). Therefore, a high density of output time points and/or a larger number of parameter sets creates the risk (due to the relatively high computational complexity of the multiplication operation compared to the addition operation) of placing high demands on computing resources. To reduce the computational complexity, you can use an alternative method, shown in Fig. 9 years By calculating the combined processing operation (for example, performing matrix multiplication) only at those moments of time when the values of the parameters change (that is, when the accepted values are applicable exactly, in I11, 121 and 12, 122), instead of interpolating the matrices A and B separately, it is possible directly interpolate the BA matrix of the combined processing operation. Thus, if the operations are represented by matrices, then each step forward in time (between time instants at which the exact values of the parameters change) will require only OM) operations (to assemble the matrices) on each set of parameters, and the reduced computational complexity will impose lower computational demands resource. Also, if the matrices A and B are such that

M» M x MUSM M). then the matrix, which represents the combined operation BA processing, will contain fewer elements than there are in separate matrices A and B together. However, the method of interpolating the BA matrix will explicitly require that both A and B be known at the same instants of time. If the points in time for which A is defined (at least partially) differ from the points in time for which B is defined, then an improved method of interpolation is required. Such an improved method according to the exemplary embodiments of the present invention is shown in Fig. 9c--9e. In conjunction with the discussion of fig. For--9e, for simplicity, it is assumed that the matrix BA of the combined processing operation is calculated as the product of separate matrices A and B, each of which was generated based on (accepted or predicted/interpolated) parameter values. In other situations, it may be equally or more preferable to calculate the operation represented by the BA matrix directly from the parameter values without going through the representation in the form of two matrix multipliers. In combination with any of the techniques presented in fig. 9c--oe, each of these approaches falls within the scope of this invention.

In fig. 9c presents a situation in which the set T1 of moments of time for the parameter that corresponds to the matrix A includes 112 time values that are absent in the set T2 (moments of time for the parameter that corresponds to the matrix B). Both matrices must be interpolated using a linear interpolation scheme and this method determines the moment (ir-Na2 of the prediction for which the value of the matrix B must be predicted (for example, using interpolation). After this value has been found, by multiplying

A and B can calculate the value of the matrix BA of the combined processing operation at the moment ir. In continuation, the method calculates the VA value at the adjacent time point 14-11, and then interpolates

VA between iz and ir. The method can also, if necessary, calculate the VA value at another adjacent point in time 14-13 and interpolate VA from yir to yiz. And although an additional matrix multiplication is required (at the moment ir-i12), the method allows you to interpolate the BA matrix of the combined processing operation directly, still reducing the computational complexity in comparison, for example, with the method presented in fig. C9a. As stated above, the combined processing operation can alternatively be calculated directly from the (accepted or predicted/interpolated) values of the parameters, rather than as a product of two matrices in explicit form, which in turn depend on the values of the corresponding parameters.

In the previous case, only the type of parameters corresponding to A had moments of time that were not among the moments of time of the type of parameters corresponding to B. In fig. 9a presents a special situation in which the time point 112 is absent in the set 12, and in which the time point 122 is absent in the set T1. If the value of BA needs to be calculated at an intermediate moment of time y" between 72 and 122, then the method can provide both the value of B at the moment yr-y2, and the value of A at the moment 143-422. After calculating the matrix BA of the combined processing operation in both moments of time VA can be interpolated to find its value at moment !". In general, the method performs matrix multiplication only at times when the values of the parameters change (that is, at times in the sets T1 and T2 when the accepted values are exactly applicable). In between, the interpolation of the combined processing operation requires only matrix additions, which have less computational complexity than multiplication.

In the above examples, it was assumed that all interpolation schemes are linear. In fig. A method of interpolation is presented in which the parameters must first be interpolated using other schemes. In this figure, the values of the parameter corresponding to the matrix A are kept piecewise constant until the moment of time M12, in which the values change sharply. If parameter values are accepted on a frame-by-frame basis, then each frame can carry signals that indicate the exact moment in time at which the accepted value is applicable. In this example, the parameter corresponding to B has assumed only the values applicable exactly at times 121 and 122. Her method can first predict the value of B at time ir immediately preceding time 112. After calculating the matrix BA of the combined processing operation at times ib and 1-1 matrix BA can be interpolated between iz and iv. Then the method can predict the value of B at the new moment of ir-H12 prediction, calculate the value of BA at the moments ir and і-4-122 and interpolate BA directly between ir and і3. And again, the operation BA of the combined processing was interpolated over the entire interval, and its value was found at all points in time of the output. In comparison with the earlier situation presented in fig. yes, where A and B would be interpolated separately and BA is calculated by multiplying A and B at each output time, fewer matrix multiplications are required and the computational complexity is reduced.

Equivalents, extensions, alternatives and more

Additional embodiments of the present disclosure will be apparent to one skilled in the art upon review of the above disclosure. Despite the fact that this description and graphic materials disclose embodiments and examples, the disclosure is not limited to these specific examples. Numerous modifications and changes are possible within the scope of this disclosure, defined by the appended claims. Any reference signs appearing in the claims should not be construed as limiting its scope.

In addition, after studying the graphic materials, the description and the attached claims of the invention, the specialist can understand and make changes to the disclosed implementation options. In the claims, the term "containing" does not exclude other elements or stages, and the singular form does not exclude the plural. The very fact that some features are mentioned in mutually exclusive claims does not indicate that a combination of these features cannot be used with advantage.

The systems and methods disclosed above may be implemented in software, hardware, hardware, or a combination thereof.

When implemented in the form of hardware, the division of tasks between functional nodes, which was discussed in the above description, does not necessarily correspond to the division into physical nodes; conversely, one physical component can perform several functions, and one task can be performed by several physical components in interaction. Some or all of the components may be implemented as software executed by a digital signal processor or microprocessor, or may be implemented as hardware or as a specialized integrated circuit. Such software may be distributed on machine-readable media, which may include computer media (or permanent media) and communication media (or temporary media). As is well known to one skilled in the art, the term "computer media" includes non-volatile and non-volatile, removable and non-removable media implemented by any method or technology to store information such as machine-readable commands, data structures, program modules or other data. Computer storage media include, but are not limited to,

RAM, SSD, SSD, flash memory or other memory technology, CD-ROMs, Universal Digital Discs (UDDs) or other optical discs for information storage, magnetic cassettes, magnetic tape, magnetic disk for information storage or other magnetic devices for storing information, or any other media that can be used to store the necessary information and that can be accessed using a computer. Additionally, as is well known to those skilled in the art, communication means typically embody machine-readable commands, data structures, program modules, or other data in a modulated data signal, such as a carrier wave or other data transmission mechanism, and include any means of information delivery.

Claims (29)

FORMULA OF THE INVENTION 1. A method of enhancing dialogue in an audio system decoder, which includes the following steps: receiving a number of down-mixing signals, which are the result of down-mixing a larger number of channels; receiving parameters to enhance dialogue, and these parameters are defined with respect to a subset of the channel array that includes channels that contain dialogue, and this subset of the channel array is downmixed into a subset of the downmix signal array; adoption of recovery parameters, which provide the possibility of parametric recovery of downmixed channels into a subset of downmixed signals; parametric up-mixing of only a subset of a series of signals; down-mixing based on restoration parameters in order to restore only a subset of a series of channels, which contains a subset of the series of channels, in relation to which the parameters for enhancing the dialogue are defined; applying dialog enhancement to a subset of the channel array for which the dialog enhancement parameters are defined, using the dialog enhancement parameters to provide at least one dialog enhanced signal; and providing dialogue-enhanced versions of a subset of the downmix signals by mixing at least one dialogue-enhanced signal with at least one other signal. 2. The method according to claim 1, which is distinguished by the fact that at the stage of parametric upmixing, only a subset of the downmix signal range is not used for the purpose of restoring only a subset of the channel range, which contains a subset of the channel range, for which the parameters for dialogue amplification are defined. C. The method of claim 1, wherein the mixing is performed according to mixing parameters that describe the contribution of at least one enhanced dialog signal to the enhanced dialog version of a subset of the downmix signal series. 4. The method according to any of the previous items, characterized in that the step of parametric up-mixing of only a subset of the down-mixing signal range includes restoring at least one additional channel in addition to the channel range for which the dialog enhancement parameters are defined, and the mixing includes mixing at least one additional channel together with at least one signal with enhanced dialogue. 5. The method according to any one of claims 1-3, which is characterized in that the step of parametric up-mixing of only a subset of the range of down-mixing signals includes restoring only a subset of the range of channels for which the parameters for dialogue enhancement are determined, and the stage of applying dialogue enhancement includes performing predicting and amplifying a dialog component from a subset of the channel array for which the dialog enhancement parameters are defined, using the dialog enhancement parameters to provide at least one dialog enhanced signal, and wherein the mixing includes mixing the at least one dialog enhanced signal with a subset of the downlink signal array mixing 6. The method according to any of the previous items, which is characterized by the fact that it additionally includes: receiving an audio signal that represents the dialog, while the step of applying the dialog enhancement includes applying the dialog enhancement to a subset of the number of channels for which the parameters for the dialog enhancement are defined, with the additional use of a beep that represents dialogue. 7. The method of any preceding claim, further comprising receiving mixing parameters for mixing the at least one enhanced dialogue signal with the at least one other signal. 8. A method according to any of the preceding claims, characterized in that it includes receiving mixing parameters that describe a down-mixing scheme that describes into which Zo the down-mixed signal is mixed for each of the number of channels. 9. The method according to claim 8, which is characterized by the fact that the down-mixing scheme changes over time. 10. The method according to any one of the previous items, which is characterized by the fact that it additionally includes receiving data that identifies a subset of the number of channels with respect to which the parameters for enhancing the dialogue are determined. 11. The method according to claim 10, and with dependence on claim 8 or claim 9, which differs in that the data identifying the subset of the channel series, in relation to which the parameters for enhancing the dialog are determined, are used together with the down-mixing scheme to find the subset of the series of signals of the down-mixing mixing, in which the downmixing of a subset of a number of channels is performed, in relation to which the parameters for enhancing the dialogue are defined. 12. The method according to any of the previous clauses, which is characterized in that the stages of upmixing only a subset of a number of signals of downmixing, applying dialogue amplification and mixing are performed as matrix operations defined, respectively, by restoration parameters, parameters for dialogue amplification and mixing parameters . 13. The method according to claim 12, which is characterized by the fact that it additionally includes combining by matrix multiplication matrix operations that correspond to the steps of up-mixing only a subset of the down-mix signal series, applying dialogue amplification and mixing into a single matrix operation before applying to a subset of the signal series reduction mixing. 14. The method according to any one of the previous items, characterized in that the dialog gain parameters and the restoration parameters are frequency dependent. 15. The method according to claim 14, which is characterized in that the parameters for enhancing the dialogue are determined with respect to the first set of frequency bands, and the recovery parameters are determined with respect to the second set of frequency bands, while the second set of frequency bands is different from the first set of frequency bands. 16. The method according to any of the previous clauses, which differs in that: the values of the parameters for strengthening the dialogue are taken repeatedly and associated with the first set of time points (T1-ШН1, M2, M3,...3), in which corresponding values are exactly applicable, and between successive moments of time it is necessary to perform the predetermined first interpolation scheme (11); and the recovery parameter values are repeatedly taken and associated with a second set of time points (T2-121, 122, 123,...3U) in which the corresponding values are exactly applicable, and between successive time points a predetermined second scheme must be performed interpolation (12), wherein the method further comprises: performing a selection of a type of parameters that are either dialogue enhancement parameters or restoration parameters, such that the set of time points associated with the selected type contains at least one prediction point, which is a point in time (p) that is not in the set associated with the unselected type; making a prediction of the value of the parameters of the unselected type at the time of (Ir) prediction; calculation based on at least one predicted value of the parameters of the unselected type and the accepted value of the parameters of the selected type of the combined processing operation, which is at least an upmixing of only a subset of the downmixing signals, followed by the amplification of the dialog at the time of (Ir) prediction; and computing based on at least the parameter value of the selected type and the parameter value of the unselected type, wherein at least one of said values is an accepted value of the specified combined processing operation at a contiguous point in time (trip) in the set associated with the selected or unselected type , and the indicated stages of upmixing only a subset of a number of downmixing signals and the application of dialog amplification are performed between the moment (p) of prediction and the adjacent moment of time (iz) using the interpolated value of the calculated combined processing operation. 17. The method according to claim 16, characterized in that the parameters of the selected type are recovery parameters. 18. The method according to claim 16 or claim 17, which is distinguished by the fact that one of the following is true: the specified combined processing operation at the adjacent moment of time (i) is calculated on the basis of the accepted value of the parameters of the selected type and the predicted value of the parameters of the unselected type; the specified combined processing operation at the adjacent moment of time (trip) is calculated on the basis of the predicted value of the parameters of the selected type and the accepted value of the parameters of the unselected type. 19. The method according to claim 16 or claim 17, which is characterized by the fact that the specified combined processing operation at an adjacent time point (time) is calculated based on the accepted value of the parameters of the selected type and the accepted value of the parameters of the unselected type. 20. The method according to any of claims 16-19, which is characterized by the fact that it additionally includes making a selection based on the first and second interpolation schemes of the combined interpolation scheme (IS) according to a predetermined selection rule, and the specified interpolation of the corresponding calculated of combined processing operations corresponds to the combined interpolation scheme. 21. The method according to claim 20, characterized in that the predetermined selection rule is defined for the case in which the first and second interpolation schemes differ. 22. The method according to claim 21, which differs in that, in response to the fact that the first interpolation scheme (11) is linear, and the second interpolation scheme (I2) is piecewise-continuous, linear interpolation is selected as the combined interpolation scheme. 23. The method according to any of claims 16-22, which is characterized by the fact that the prediction of the value of the parameters of the unselected type at the moment (Ir) of the prediction is performed according to the interpolation scheme for the parameters of the unselected type. 24. The method according to any one of claims 16-23, characterized in that the combined processing operation is calculated as a single matrix operation before applying it to a subset of the downmix signal series. 25. The method according to claim 24, which differs in that: linear interpolation is chosen as the combined interpolation scheme; and the interpolated value of the corresponding calculated combined processing operations is calculated using linear matrix interpolation. 26. The method according to any one of claims 16-25, which is characterized in that said received downmix signals are divided into time frames, and in a fixed mode of operation, the method includes accepting at least one value of parameters of the corresponding types, exactly applicable at the time in each time frame 27. A method according to any of the preceding claims, characterized in that the mixing of the at least one dialogue-enhanced signal with the at least one other signal is limited to an incomplete sampling of the downmix signal range. 28. A computer-readable medium containing instructions for performing the method according to any one of claims 1-27. 29. A decoder for enhancing dialog in an audio system, which includes: a receiving component configured to receive: a number of down-mixing signals, which are the result of down-mixing a larger number of channels, parameters for enhancing dialogue, and these parameters are defined with respect to a subset of the number of channels, which includes channels that contain dialogue, and this subset of the number of channels is down-mixed into a subset of the down-mix signals, and restoration parameters that enable parametric restoration of the down-mixed channels into a subset of the down-mix signals; an upmixing component, made with the possibility of parametric upmixing of only a subset of the downmix signal series based on restoration parameters in order to restore only a subset of the channel series, which contains a subset of the channel series, in relation to which the parameters for dialogue amplification are defined; and a dialog enhancement component configured to apply dialog enhancement to a subset of a number of channels for which dialog enhancement parameters are defined, using dialog enhancement parameters to provide at least one dialog enhanced signal; and a mixing component configured to provide dialogue-enhanced versions of a subset of a number of downmixed signals by mixing at least one dialogue-enhanced signal with at least one other signal. r 100a pe z E i y Me: ey C) 2 G v 1RE lk ce K te /z la E tvoje / tva ti M/o iv kv sho yav Fig. And V. 1006 2 v/s w d: Y t I. " iv Y e TSO St vv ! Skeleton vh (vv Fig. l Short circuit! | -ViZH V 'OO 8IV ; and | and and! : пн PI U Ж Ж Z Esh zh I roi p: Tj Z reno reeheya I fen vo E | Well, OYAZE! ! Ж rev rynnnnnovv "I AND N 1 SAME OK N gozhe neo I So I reg I i ke VU SCHE rooj iOTEya: TOO TOZ i fen nd nyu m хх хм сх pf V, NU I I 5 Her ze I and ze - zh ! их ии : а : гы : N eat rhymes pi fen v : Eh yiv zvi - OT I :OTVE N : : ж H 3 : : iv N I I xx v VN A NN : o E pi what i unchchnyannpolnyaofvh v U j zha tyaotyay nyah nn y E Shchi y I I a - Fig. 2 cha y xx Y UK k ku Ei M kRA-k eat Sh ih ih . In V rn Krh I» i-i dk yah, sh g p --k sok Fig. From drinking; n I - | Pon ov I VO : shi chn aa t i t v djan N N x N chiny : N ' N x N : : ta i N | zn Fin ' S : : V U INN butttntty N : : tu NI N BO 2 fo- MNK nn t sn: Kolya SHE zh N : i TT, Her on nn non z V nn ANU Fig. 4 che V I tu fitttnttntntnnnn 7 ZE still N i i I: --e ож ж yuyu syu ожюж o зжжо ож жжя ж LO ЖЖ Ж Z SYLY ЖХ ЖХ ЖХ ХАМО МАХ ЯМ ш шю жлюо ож шжж Х UYUYU жу жжх зл С ж. AH, II av: ! N also I I a : ZI 53 : 1 N dok dk x, x i i zhu Ku t I N N ie pana i p a a Z vi mielyolh liliya. tin, : CASH and : : rekh spend duukhkyyuukhkhkhkhhtkhkhkhkhkkkhhya reko yeyukkeeekn Zh : ut and ! ! ! me: oh and NU! N N Oh N ov M is : and : | kind of N x ! N y N N N . : x : : SIM NU oo: i i i M i : : but ! 5 th! : N he NU e div a Shchi zh | also: DE che i oj Ge: rezh tk N ' sh NN N iyi; rock N HY : ! and This and N NE ii 5 and id Z : and I oon Z Uh R I BM o ! and shock of ZK SHU : SE : x : OO Zhoozk N | х рони Вк Ше TV ж Uetnnnnnn nnya fic eat where oh : ! kN: min NOT in ish I | i v, o kA g MENE NI SN pz MI i V; se» am ssh ni OT OE Mi ra i i ZY; k izhe V bita: K- and vk : y shov and m I : OK obo. tok h i 1 x ' ih "k Z seven y shi i nn Not Zh. r ! | Kk sh m shchi her frebeskissfie 10 ee and Her same as a Ach oz Te jealously: I and zhl : and OV and and ka N : M lllyallyalyaeyyunyuyuyuyuyuyuyuyukyutyky Mody V ddhllzhzhzhhllllllllllalALAHALAM ; po I J lo mk mk io snnh sil te vntyh, AK OMZHHO OM OHHO CHAK RKO M MR zh zhk zhk zhk skho skh mk MA AJH Ah MA Zhe tk zzhjonyu Y ! iii m 1 ze ME: inn nn a en ooo pn ti: I AND N V! SHE: Mon no oo he nn nn mn I : : I : : : ; and ! i N NU pe I chy nn p SHE ve N N i I you Fig. 5 and Z «ee E Bib : : m pennnnan tya AKAAA KAH AHA TAKA KAK tt ndlyalnyalnh plAAAAAAAAAN i se E i j : i i VSH: ! "vya EE acheh she Z Ko | and vla BOaV STILL; and H; and : From ha. | ke x ; and with N M y y still y V Sya yah y - N ti EI Х donos ono muh y p py check ky re y y y Y vennnnnnu V y EI, her; N Shen gi i I ri : shi : y i ko i! | game and what | ! : k Z 1 N KE t ik ' | Y x i y | KR dumu zhov no x sei Sh- o o give AVnvIuvyutie t 5 nenny Yanizhuvalne DE K- o i ! Sya MiyuUvanja 7 a NI TV mixing Mi Eomnnnny -i Mn 7 i 1 same e : Kik t NI Vdetu s : : 1 still NO: : ; » oyu ts V i ; | 7 dates: shooi ee: Cake in ikAAAAAAAAAAAM, ! more V ro Ko iE | sa Pavniuvalne tri ov puru ryu uvaNe AND TOZ same Ts Hv Z fs mishnunia PA ! PAK odkruuvanyaya 11 Y NE NN AKA veins I ore " I; N i Z: So the same lower N and Bones N: sHE BLINDED MYT tires still Pon shi of - y k che Sh- shk SHE KO ia EE yav seborrhea sh yiv as : VAG : She | : same pt KENE U I I to teki of i : I : : fe oon nn nn nn and what : U r Y zu - ge keshu SO With sokkkyueyuyuyuyuyuuki and Her "xx stru : ! xx; i Yu I 7 o nn nn nn rya : ' sec ozh yi kh i | "7 16 : J I | Ki I : fen tu Yi i i i i x : | Ж Зі і : і Ж і 1 : Ж і 3 7 і че s chuezhim і і З N vyh na і (in ZK KU IV і і ДН y xx : х Ж х гі ниня ТЯ тва 0006 : Н : і N Y KI I : Н N В х Н Н х и ! ke ti N y feyukkkkkkkni N : t sche N pn N H duy i N : ! i NN PTK Ri a i : ' ORI Ba E ish ! : N S zhe MNN i N ' : ' ZY i SU : i x Z N : ShK Ka: and write in M shi ri !' | M Ps pnnvnnnn: NAKYASHCHE BO. vat More KOR aa PABNSHSHCHUVVLN Z GO chay she TT : t : 12 : hOETYA : sm tone MI MNN ; sho N YAM Ro ren MIK NaNVYA both a rod pk, and E | V shih pa "Vyn NI i, : ; shin nn vh U NE and s shin a she pol Z OZHYASH R svertttttt ts an khz: Z VE, nb r z yi zhi U nen in GSK Y Biju vk in I Ya ki R od KLVNEDUVEKNE tk) b37; dis NK NE Nes I g her : "5 NU min : i VE So rbenbeeuernvneyayavya MiIBNEVINYNA Hor. sha : same vm and EE Ku t. r: | : Z shen HOR mozhn sho zhi : ! Sa na p V MO, SCHE Sov : : TOV Moeeetstekin sa N Y oh as "KK, : : EM i N nn i a KI s Zk. damn it! nothing | и ши пом ке пн H 1 ЖК... и и т и и и Теж У м : еще dlllknlnnnnn фи и ом и : | o z : V Yake klaaAAAAAH AHA KANA NAAN TRUCK oi same i I B : U N : oto i rent t Sy Chi: : shchi i : pe Du le N zh TEO. : Yes! : it's your fault N y oh V ! : sho i k POL i I Ka H i - zy in fentnnnnnnnnnnnnnnntn nnya ! th food VE PAAAAAAHACT tt tt tt TA A КААААААААКАЙК kn» ше | ! : Kiy a Ve N p p ia : zha NK ; ! : yin pen Mo oddnnntnttttttttttttnth : ka pdddetenennnya p Kk tk tt t U oo zhilzhizhko VD» : i zh x : : 7 x N N TI U Her oh z Z 1 Fig. 7 taz i es hu y as I ze s N schi I -y and i. -KaO SHE: i tu smy U V y go dya a t E o : i Kt in Doezhkh i N ' Kotseannya SHE z : Y N ee N 2 rev Z KEFEUETU Y tt ddddddlstnya y y ! and: Ken zennnia Dila! Sha She gave N N ki i : EV i : Esh: i ag i seoshmk y pittya apa AKAHAHAYU tt tt tt VK ev VIV, 1 more SHE ye same : ii Zinzhuraz Okh N MTA KU v, ri: : : u o Yanyazhuvalne V n KI me lyko i AV idauchyayo : 7 N : i Z aukakytsvto : MK fev nin i In the fountain MNaremetrnho | Also: NO and YES: ; х Н Н : : A A and Her defect toll E g. ! and Coding; p IN reco t FROM them. 2nd Hek nanni dov in her university N N Z : ! AND; EN U. date alla shcha Y psnkny Byy g ZY o NO I UA St h ne M nan ny pellage vv tUlttkvaeyuk vo breathing yannia zeshkzhzhzhzhe zhk z iro yayu zasut I u Fig. Cha A. Va mini nya an V V ov she V No Me Fig. or Va pd saw shihlskhtuyueuuat rnyh rnyh ktaya uh: g y yakkzhezhvekhhya skkhath hori TK fig. fri Az dechhlnya S and M ' pt at 2 PZ Z Fig. By And with technicians and zhtynnh nanny vv! there is no womb Anna And in : ! and Gu E: Ma sko u pyhoyalyaniya aa o zhlgoltekheuuyu kyazhuuyekh hv KZ i iz I Fig. Ze